I took some recent times at 60W and 76W and made some power trend plots of the first three hours of the lock. The 60W time is from a long lock around March 29th, and the 76W time is from a long lock around April 29th. Unfortunately the time for the April 29th lock we had the POP beam diverter closed, so we are missing the sideband trend during thermalization.
I found a shorter lock at 76W on April 20th where the POP beam diverter was open. Trends attached.
I have attached another set up plots, this time adding in a 60W lock trend from early January.
One major effect we have considered when looking at these plots is the trend of the 9 MHz sideband as the IFO thermalizes. If you look at the first plot in both the main alog and the two comments, the bottom right corner has two plots showing the POPAIR RF18 and RF90 trends, effectively the 9 and 45 MHz buildups. When we lose our 9 MHz sideband, PRCL becomes unstable. Currently, the thermalization guardian works to correct exactly this, by increasing the PRCL digital gain to maintain a ~30 Hz UGF. We have mainly seen this as an issue with operating at 75 W. However, these power trends are showing us that the 9 MHz loss is present at 60W too, and Evan and Dan held the IFO at 25W a few months ago and demonstrated that this loss occurs even at low power.
The new plot I have added is an interesting comparison. On the POPAIR RF18 plot, you see that the blue trace- 76W lock from April 20, shows that we power up and then lose significant 9 MHz power. The red trace shows something similar from Jan 6 in a 60W lock. We also see this loss in a March 60W lock. I included the red trace to compare with the green trace because in the green trace we stopped at 25W for long enough to lose some 9 MHz, and then continued up in power. The green trace is also after we have improved the mode matching with new ring heater settings. It appears that build-up wise, our current situation is very similar to where we were in January.
However, this is further proof that we see 9 MHz loss no matter which operating power. I will look for more locks to compare and see if the difference between the red and green traces at 60W is a result of the TCS change. One thing to note in particular is that the ring heater changes reduces the build ups, so from January in red to March in green, we reduced the power recycling gain at 60W.
